The tone and language used by Professor Rhys Kesselman in his criticism of our recent study on the CPP and RRSPs is a far cry from the professionalism and decorum one would expect from a distinguished economist and holder of a Canada Research Chair. Rather than stick to debating the facts, Professor Kesselman stoops to innuendoes and school yard taunts to disparage our work. Reasonable people should be able to disagree without being disagreeable.
To start, it is important to note that our study was formally reviewed by two academic economists. Kesselman’s approach to his disagreement with our study is reflective of a larger problem in Canadian universities. When a student’s work disagrees with the views of their professor, their chances of success are much lower. This is not what applied policy research is supposed to be about.
The central point of our study is that proponents of an expanded CPP, which Kesselman clearly is, often ignore or downplay people’s decision-making about voluntary savings like RRSPs when government forces them to save more through the CPP. But the economics taught to students by instructors like Kesselman teaches them that individuals make decisions about their spending and savings based on their preferences and expectations of long-term earnings. This theory, which is found in virtually all economic textbooks, has proven robust over time as countless researchers have applied and tested it. The fundamental idea is that people will reduce their voluntary savings in response to an increase in mandatory savings if their preferences and expectations for lifetime income do not change.
Nowhere in Kesselman’s criticism does he dispute this basic economic concept. But it is this idea that has been lost in the broader debate about expanding the CPP. If people actually save less voluntarily in a way that offsets increased mandatory savings, overlooking this behavioural response runs the risk of inflating the benefits of a compulsory expansion of the CPP.
Putting aside Kesselman’s disparaging remarks, he does raise some worthwhile questions.
We completely agree that a more detailed analysis of micro data incorporating a host of potential explanatory variables is needed. That’s why we made this precise recommendation in our study and clearly stated that our analysis was preliminary based on readily available data from the Canada Revenue Agency.
Kesselman mistakenly argues that historical CPP changes are not comparable with current proposals for an expanded CPP. His rationale is that the former entailed CPP contribution increases to improve the solvency of the plan without any benefit increases, whereas current proposals for CPP expansion include increases to both contributions and benefits.
This argument does not account for the expectations of Canadians regarding CPP benefits. Opinion surveys prior to 1997 repeatedly showed that Canadians, particularly younger and middle-aged workers, did not expect to receive CPP benefits (or at least full benefits) due to the deficiency between the resources available and benefits promised.
To borrow the terminology of Nobel laureate Robert Lucas, Canadian workers did not have a “rational expectation” of full CPP benefits under the pre-1997 system. Expectations of receiving promised CPP benefits increased after the reforms as the program moved to a more sound financial footing. From this vantage point, Canadian workers did experience an increase in their expected benefits. To say there is “total incomparability between the events” is incorrect.
Kesselman then oddly challenges our findings by using the same methodology, which he criticizes, to analyze the 2004 to 2008 period. At the very least, such an approach is inconsistent. And critically, this period includes 2008 data, which as all Canadians are painfully aware was the start of the recession and thus greatly the exaggerates the declines in RRSP contributions noted by Kesselman (in one case the percentage point decline is reduced by more than half when 2008 data is excluded).
More troubling is that he presents the 2004 to 2008 data as if it definitely disproves the preliminary results we reported in our study. They don’t. What they do is raise additional questions about other factors influencing RRSP contributions, which again we thoroughly agree with.
All this reinforces the main point of our paper, which seems lost on Professor Kesselman. A more detailed analysis of individual tax returns controlling for a host of potential explanations for the decline in RRSPs is warranted given calls to increase the CPP. Such an analysis would more definitely ascertain the extent to which Canadians are expected to offset increases in CPP with lower RRSP savings.
Notwithstanding Kesselman’s longstanding advocacy for an expanded CPP, a broad, genuine assessment of the costs and benefits of this proposed reform is still required.